Experts urge investment in local health data, Nigerian-owned AI models

By Ojoma Akor

The Founder and Chief Executive Officer of the Africa Hub for Innovation & Development (AHFID), Dr Kunle Kakanfo, has called for greater investment in locally generated health data and Nigerian-owned artificial intelligence (AI) models to ensure that AI technologies deliver accurate and relevant health outcomes for Nigerians.

Kakanfo said Nigeria risked becoming overly dependent on foreign AI technologies if it failed to invest in collecting, structuring and securing local health data that could be used to train models designed for the Nigerian population.

He made the call ahead of the forthcoming AI in Health Summit in Abuja, where data sovereignty and locally trained AI models are expected to feature prominently.

According to him, many AI models used globally are trained largely on foreign datasets, raising concerns about their accuracy and suitability for African populations.

“One of the things, or one of the conversations, will be shining on how to have locally trained models using our local data to be able to train models that can help us better predict outcomes for people,” he said.

Kakanfo said Nigeria should move beyond importing AI tools to building models that understand the country’s unique population, diseases, languages, health behaviors and clinical realities.

He cited ongoing work in hospitals in the Federal Capital Territory, where researchers are collecting cough sounds to explore developing AI models that can detect tuberculosis.

According to him, cough sounds could potentially identify other health conditions, while research in other settings has also explored infant cries as a potential signal for detecting neonatal sepsis.

Kakanfo highlighted that Nigeria already has enormous amounts of potentially valuable health data but lacks sufficient systems and investment to turn it into usable intelligence.

“Data is the foundation,” he said, stressing that no single institution could independently create all the structured data required to develop effective AI models.

He said government, which holds a significant proportion of the country’s health data, has an important role in creating an ecosystem where data can be responsibly curated, protected, and made available for research and AI development.

However, he noted that the responsibility does not rest with government alone, as Nigeria also needs the technical capacity to convert raw data into useful intelligence.

“If that data cannot be used to create models that can improve the health outcomes of the people, then we’re limited in how much we could explore the use of artificial intelligence,” he said.

‘Local data could improve diagnostic accuracy’

Kakanfo said locally trained AI models could improve the accuracy of diagnostic tools because they would be developed using data that reflects the population in which the tools would ultimately be deployed.

He pointed to a locally developed cervical cancer screening tool that uses images of the cervix to identify precancerous lesions, describing it as an example of how locally generated data could support AI applications tailored to Nigerian healthcare needs.

He said similar approaches could be explored across disease detection, health surveillance, clinical decision support, supply chains and other areas of healthcare.

The AHFID chief executive also warned against unquestioningly procuring foreign AI systems without assessing their performance in the Nigerian context.

He said imported tools may work in Nigeria but may not deliver their best possible performance because the underlying models may have been trained on populations and datasets that differ significantly from those in Africa.

NCDC backs local innovation

Senior Technical Adviser to the NCDC Director-General, Dr Abiola Oshunniyi, similarly called for stronger pathways to enable Nigerian health technology innovators to have their solutions validated, adopted and scaled.

He said Nigeria has several health technology solutions that could be deployed more widely if appropriate standards, regulatory pathways and adoption mechanisms were established.

“Why are we bringing solutions from outside the country when they can bring local solutions that have international standards and can even be exported across Africa?” Oshunniyi said.

He said Nigeria should create mechanisms through which a solution developed by a young innovator in a city such as Lokoja could potentially be adopted and scaled by government.

Oshunniyi stressed that regulation, standards, and adoption must work together to create an enabling environment for local innovation.